Field
This invention relates generally to data communications, and more specifically, to a method and system to service load balancers.
Background
Web services and cloud computing are deployed in an unprecedented pace. New servers are unloaded and installed at datacenters every day. Demands of web services and corporate computing come from all directions. Consumer oriented services include iPhone™ apps, mobile applications such as location based services, turn-by-turn navigation services, e-book services such as Kindle™, video applications such as YouTube™ or Hulu™, music applications such as Pandora™ or iTunes™, Internet television services such as Netflix™, and many other fast growing consumer Web services. On the corporate front, cloud computing based services such as Google™ docs, Microsoft™ Office Live and Sharepoint™ software, Salesforce.com™'s on-line software services, tele-presence and web conferencing services, and many other corporate cloud computing services.
As a result more and more servers are deployed to accommodate the increasing computing needs. Traditionally these servers are managed by server load balancers (SLB). SLB are typically network appliances, such as A10 Network's AX-Series traffic managers. SLB manage the load balancing of servers based on incoming service requests. Common methods to balance load among servers is to distribute the service requests based on the applications (HTTP, FTP, etc.), service addresses such as URL, priorities based on network interfaces or host IP addresses. SLB may distribute service requests additionally in a round robin fashion to the servers, assuming and ensuring the servers would be evenly loaded. However, different service requests have different service computing consequences. A server may be fully loaded with only a handful of service requests while another server remains mostly idle even with plenty of service requests. SLB may inappropriately send another request to a busy server, incorrectly considering the busy server being readily available, instead of sending the request to an idle server.
It would be beneficial if the SLB are aware of the computing load situation of a server so that SLB can better select a server to process a service request.
Therefore, there is a need for a system and method for a server load balancer to select a server based on the server load status.